WEBVTT 00:02.680 --> 00:06.260 Hi, good afternoon, and welcome to today's IPAT 00:06.260 --> 00:08.880 -GVU lunch lecture. My name is Michael Best. I'm 00:08.880 --> 00:11.750 executive director of the Institute for People and 00:11.750 --> 00:14.550 Technology, and I'm really glad to be hosting Professor 00:14.550 --> 00:18.060 Susan Weish this afternoon for a talk that I'm 00:18.060 --> 00:20.610 sure is going to be really, really interesting. 00:21.720 --> 00:26.700 Next week, lunch lecture, is Haoqi Zhang, 00:26.700 --> 00:30.050 who will be presenting on computational ecosystems. 00:31.900 --> 00:33.640 Ecosystems or ecosystems? 00:33.640 --> 00:35.450 I generally go with ecosystems. 00:35.670 --> 00:37.780 Let's have, by a show of hands, 00:38.200 --> 00:40.940 how many people go with ecosystems? 00:42.500 --> 00:43.620 Okay. 00:46.640 --> 00:50.220 Computational ecosystems, advancing human values 00:50.220 --> 00:54.360 through integrative computing and changing practice. 00:54.360 --> 00:58.260 So please do come next week, same time, same place. 00:59.040 --> 01:02.420 Does anyone need to leave at 1.20 01:02.420 --> 01:05.880 because you have to be somewhere at 1.30? 01:07.880 --> 01:08.880 You. 01:10.760 --> 01:11.760 You. 01:11.940 --> 01:15.820 So at 1.20 we'll take a beat and please do leave but 01:15.820 --> 01:19.360 otherwise the presentation with Q&A will go to 1 01:19.360 --> 01:24.160 .30 and as a matter of respect for the speaker and the 01:24.160 --> 01:28.120 rest of us don't be please getting up and loudly. 01:28.760 --> 01:31.420 exiting in the middle of Q&A, 01:31.660 --> 01:32.660 capisce? 01:32.860 --> 01:37.440 Okay, it's really my great pleasure to introduce 01:37.440 --> 01:42.060 Dr. Susan Weish, a colleague, a friend who I've known 01:42.060 --> 01:45.500 since she was, a PhD student here at Georgia Tech. 01:45.500 --> 01:48.280 We were just talking and I think we maybe overlapped 01:48.280 --> 01:53.260 about four or so years here at Georgia Tech 01:53.260 --> 01:56.940 and I always remembered two things. One is that she 01:56.940 --> 01:59.340 was doing work that I thought was just sort of one 01:59.340 --> 02:02.240 of those, wait, is that allowed? Can you do this? 02:02.820 --> 02:06.120 Studies of digital spiritualism and religion. 02:06.240 --> 02:09.780 And the other thing is that Susan's a really good 02:09.780 --> 02:13.200 designer. You're a good designer. And so she 02:13.200 --> 02:18.000 would do some sort of exploratory kind of HCI, 02:18.000 --> 02:22.200 human-centered design activity, but everything would 02:22.200 --> 02:25.400 just look great, like totally professionally 02:25.400 --> 02:30.380 done. And I was so jealous of that design skill. 02:30.580 --> 02:35.500 After graduating from Georgia Tech, 02:36.100 --> 02:38.860 she went on to Michigan State University, where she 02:38.860 --> 02:41.640 is currently associate professor in the Department of 02:41.640 --> 02:45.040 Media and Information and a faculty affiliate of their 02:45.040 --> 02:48.420 university's African Studies Center. Her research 02:48.420 --> 02:53.220 focuses on HCI and ICTD, and for nearly two decades, 02:53.220 --> 02:57.100 she sought to understand how people in Africa use ICTs 02:57.100 --> 03:00.120 and design technologies that reflect their contexts, 03:00.320 --> 03:02.840 critically examine whether ICTs support 03:02.840 --> 03:06.300 socioeconomic development, and increase African 03:06.300 --> 03:09.500 participation in technology design and research. 03:09.500 --> 03:12.400 And to manage these questions, she's conducted 03:12.400 --> 03:16.000 extensive field work in Kenya since 2007. 03:16.440 --> 03:18.880 And so there's two other things I'd like 03:18.880 --> 03:21.040 to mention about Susan's scholarship. 03:21.040 --> 03:26.140 One is that it's this consistent investment 03:26.140 --> 03:29.900 in Kenya and in relationships and in 03:29.900 --> 03:32.720 communities that I really admire in the work. 03:32.720 --> 03:36.620 That she's put in now 20 years, basically, 03:36.860 --> 03:38.900 19 years, I guess. 03:39.300 --> 03:41.860 In the same part of the world, 03:41.860 --> 03:43.820 working on these questions, 03:43.960 --> 03:45.580 building trust, 03:46.060 --> 03:47.440 building understanding, 03:47.760 --> 03:51.320 learning Ki Swahili, and that kind of 03:51.320 --> 03:53.500 commitment, I think, is just so admirable. 03:54.020 --> 03:56.700 A second thing that I find very interesting, 03:56.700 --> 03:59.400 and I wonder if we're going to have any taste 03:59.760 --> 04:02.760 of this today, is you have these kind of 04:02.760 --> 04:06.020 questions. You often answer them in the negative. 04:06.460 --> 04:10.090 So you might be saying, like, will this agriculture 04:10.680 --> 04:15.100 information system help these folks in some 04:15.100 --> 04:19.100 place in Kenya, and sometimes the setup kind of 04:19.100 --> 04:21.620 leads you to think that the end result is going 04:21.620 --> 04:24.740 to be, yes, of course, it's great, but I think 04:24.740 --> 04:28.880 in your scholarship, your point is not so much. 04:28.880 --> 04:31.760 So I think that that's really an interesting part 04:31.760 --> 04:35.320 of your work. So, please, without further ado, 04:35.480 --> 04:36.740 Dr. Weitsch. 04:36.940 --> 04:37.820 Thank you, Mike. 04:41.090 --> 04:42.520 So, thank you, Mike, for that 04:42.520 --> 04:44.220 very generous introduction. 04:44.220 --> 04:46.240 I'm Susan Weitsch. I'm an associate 04:46.240 --> 04:48.260 professor in the Department of Media and 04:48.260 --> 04:50.660 Information at Michigan State University. 04:50.740 --> 04:53.760 It's wonderful to be back at Georgia Tech. I want to 04:53.760 --> 04:56.880 thank Priscilla for all the work that she's done to host 04:56.880 --> 04:59.800 me today. And today I'm going to describe how 04:59.800 --> 05:03.720 accompaniment or an approach to research that broadly 05:03.720 --> 05:09.160 prioritizes relationships over research outcomes can benefit 05:09.160 --> 05:12.000 human-centered approaches to technology development. 05:12.200 --> 05:17.740 And I fully realize that Carl DeSalvo gave a talk 05:17.740 --> 05:21.160 about this same topic just last semester. And it's 05:21.160 --> 05:24.540 a genuine, a truly genuine coincidence that we've 05:24.540 --> 05:28.410 both arrived at this idea and how it might apply 05:28.440 --> 05:31.140 to design in our work. I'd like to think that 05:31.140 --> 05:33.920 my work builds on Carl's ideas, like Mike said, 05:33.920 --> 05:37.220 drawing from my long-term fieldwork in Kenya and 05:37.220 --> 05:39.720 bringing a sort of global perspective to this topic. 05:39.970 --> 05:42.840 At the end of this talk, I hope you'll appreciate 05:42.840 --> 05:45.680 how human-centered design can be used to 05:45.680 --> 05:49.080 support longer-lasting and more meaningful outcomes 05:49.080 --> 05:52.180 that aren't necessarily digital technologies. 05:54.860 --> 05:58.480 So I'll begin with an outline for my talk. I'll tell 05:58.480 --> 06:01.260 you a little bit more about me and my research. I'll 06:01.260 --> 06:04.540 give a very brief explanation of human-centered design, 06:04.540 --> 06:08.220 this popular approach to technology development. and 06:08.220 --> 06:11.300 I'll primarily focus on critiques of this approach. 06:11.740 --> 06:15.040 And then I'll argue that accompaniment offers 06:15.040 --> 06:17.760 a way to address these critiques by drawing 06:17.760 --> 06:19.740 attention to different outcomes that are 06:19.740 --> 06:22.600 not necessarily technological interventions. 06:23.150 --> 06:26.150 And to do this, I'll talk about two projects. 06:26.360 --> 06:29.420 The first is where we used human-centered design 06:29.420 --> 06:32.080 to collaborate with Kenyan youth to develop 06:32.080 --> 06:35.140 an intervention to support type 1 diabetes 06:35.140 --> 06:39.780 self-management. The second project is another 06:39.780 --> 06:41.960 one where human-centered design was central. 06:42.060 --> 06:45.280 It's part of the Mozilla Responsible Computing 06:45.280 --> 06:48.000 Program. This is a program that supported 06:48.000 --> 06:51.360 the redesign and implementation of an HCI course 06:51.360 --> 06:54.580 so that it centers Africans' perspectives. 06:55.020 --> 06:57.140 This is a course that I developed and have 06:57.140 --> 06:59.440 taught at Michigan State where students 06:59.440 --> 07:02.260 collaborate with their peers from Kenya's Egerton 07:02.260 --> 07:04.200 University on a design project. project. 07:04.400 --> 07:07.800 I'll discuss what connects both of these projects. 07:07.800 --> 07:10.640 I'll talk about how accompaniment can reshape 07:10.640 --> 07:13.140 what we think about as outcomes and design, 07:13.140 --> 07:15.980 and the broader implications for research and 07:15.980 --> 07:18.560 also how we teach human computer interaction. 07:20.340 --> 07:23.320 So a little bit more about me. I mean, since 07:23.320 --> 07:25.060 I've graduated, I've continued to work 07:25.060 --> 07:26.990 in the field of human computer interaction. 07:27.340 --> 07:30.280 Also, I know Mike from working in the field of 07:30.280 --> 07:33.180 information, communication, technology, and development, 07:33.300 --> 07:37.000 a field that's broadly defined as how technologies 07:37.000 --> 07:39.940 can be appropriately designed to address 07:39.940 --> 07:43.060 users' needs and to support socioeconomic 07:43.060 --> 07:45.740 development in so-called developing regions. 07:46.280 --> 07:48.460 And, you know, I think Mike can attest to this 07:48.460 --> 07:50.420 too. Over the course of our careers, it's been 07:50.420 --> 07:53.300 really exciting to see growth in these studies that 07:53.300 --> 07:56.100 could be characterized as ICTD in our field. 07:56.100 --> 07:59.620 They're incredibly valuable for broadening researchers' 07:59.620 --> 08:02.600 knowledge of how to design technology for 08:02.600 --> 08:04.860 contexts that are outside of North America and 08:04.860 --> 08:07.820 Europe or those more commonly studied in the field. 08:08.180 --> 08:12.060 And really useful for providing us with a global 08:12.060 --> 08:14.760 understanding of technology design and use. 08:19.170 --> 08:20.980 I don't want to knock this over. 08:21.370 --> 08:23.520 And like Mike said, my work has 08:23.520 --> 08:25.660 contributed to these fields in these ways. 08:25.920 --> 08:28.440 I've done a number of qualitative studies 08:28.440 --> 08:31.360 examining mobile phones, social media use, even 08:31.360 --> 08:35.100 agricultural hand tools, mostly in Africa, and 08:35.100 --> 08:37.600 use findings from these studies to offer really 08:37.600 --> 08:39.940 practical guidance on how to improve the 08:39.940 --> 08:42.660 design of these technologies for this context. 08:43.000 --> 08:46.440 I also use my findings to raise really critical 08:46.440 --> 08:48.760 questions about whether or not technology can 08:48.760 --> 08:52.340 actually support socioeconomic development, and 08:52.340 --> 08:54.740 I've also been deeply committed to diversifying 08:54.740 --> 08:57.840 perspectives within computing and design fields 08:57.840 --> 09:00.800 by increasing African participation in them. 09:02.700 --> 09:05.800 To do this work, I almost entirely conduct 09:05.800 --> 09:09.460 fieldwork in Kenya and East Africa, primarily in the 09:09.460 --> 09:11.900 rural areas, like you see in the picture here. 09:11.960 --> 09:15.920 I first began working in Kenya as a PhD student here 09:15.920 --> 09:18.860 at Georgia Tech, and since then I've been returning 09:18.860 --> 09:21.960 regularly. I typically travel there twice a year. 09:21.960 --> 09:26.020 Over time, I've developed an intermediate proficiency 09:26.020 --> 09:29.900 in speaking Kiswahili. I'm also very fortunate to have 09:29.900 --> 09:33.840 a very talented instructor and like Mike said and 09:33.840 --> 09:38.860 I think this is worth saying again the sustained long 09:38.860 --> 09:41.320 -term engagement with the people in places in the 09:41.320 --> 09:45.820 country has deeply benefited my research it's allowed 09:45.820 --> 09:49.480 me to build trust develop meaningful collaborations 09:49.480 --> 09:52.680 and gain a much deeper understanding of this context 09:52.680 --> 09:56.440 and I ultimately believe these long-term commitment, 09:56.440 --> 09:59.200 has helped me conduct research that's more ethical, 09:59.380 --> 10:02.320 that's more grounded and responsive 10:04.200 --> 10:06.520 to the communities I work with. 10:12.220 --> 10:15.480 Also happens to be, it has been, and it continues 10:15.480 --> 10:18.380 to be an exciting place to study technology. 10:18.380 --> 10:20.400 It's a place that's very much at 10:20.400 --> 10:22.380 the forefront of technology use. 10:22.500 --> 10:26.460 From the development and use of mobile money systems, 10:26.460 --> 10:29.780 notably M-PMASE, that are fully integrated into everyday 10:29.780 --> 10:33.820 life there, to using WhatsApp to share crop prices 10:33.820 --> 10:37.080 and organize digital Harambes, which are these community 10:37.080 --> 10:41.100 fundraising events where everybody chips in money. 10:41.120 --> 10:43.700 And it's these innovations that have emerged 10:43.700 --> 10:45.800 from this context that eventually find 10:45.800 --> 10:48.740 their ways to other parts of the world. 10:50.830 --> 10:53.640 And over the course of my career, I really adopted 10:53.640 --> 10:57.280 this radically interdisciplinary approach to research. 10:57.760 --> 11:00.800 I've worked closely with students and scholars 11:00.800 --> 11:03.600 from geography, anthropology, computer 11:03.600 --> 11:07.440 science, linguistics, even soil science in Kenya 11:07.440 --> 11:09.600 and at Michigan State to do this research. 11:09.600 --> 11:13.660 And a few projects. I've developed educational 11:13.660 --> 11:17.040 videos that help rural women in western Kenya 11:17.040 --> 11:20.250 learn how to use their mobile phones to access 11:20.250 --> 11:22.880 health information and agricultural information. 11:23.200 --> 11:26.440 I've worked with smallholder farmers also in 11:26.440 --> 11:29.540 Western Kenya to redesign the agricultural hand tools 11:29.540 --> 11:32.690 they use. This is called a djembe in Kiswahili. 11:33.160 --> 11:35.920 Finally, working with students, I've designed 11:35.920 --> 11:38.320 and evaluated systems in these contexts. 11:38.420 --> 11:42.280 This is a system called ImKulinda. It's a domestic 11:42.280 --> 11:46.160 security system designed for rural African households. 11:46.160 --> 11:48.260 You can see from the picture here that 11:48.260 --> 11:50.480 it's solar powered. many of these homes are 11:50.480 --> 11:53.000 off their country's electricity grids. 11:53.160 --> 11:55.900 There's a sensor based in there. When the sensor 11:55.900 --> 11:58.860 field is broken, an alert is sent to the user 11:58.860 --> 12:01.000 letting them know that somebody has come into their 12:01.000 --> 12:04.600 compound. And we've evaluated these and other sensor 12:04.600 --> 12:07.800 based systems at homes in Kenya and in Malawi. 12:09.580 --> 12:12.200 So these projects, like many in the 12:12.200 --> 12:15.440 HCI and ICTD fields, has been guided by 12:15.440 --> 12:17.440 this human-centered design approach. 12:18.160 --> 12:20.480 I'm not gonna spend too much time explaining 12:20.480 --> 12:22.900 this to this audience, but it's an approach 12:22.900 --> 12:24.800 to technology development that involves 12:24.800 --> 12:26.700 putting people, their context, and their 12:26.700 --> 12:29.050 activities at the center of the design process. 12:29.280 --> 12:31.160 It generally has these three 12:31.160 --> 12:32.940 phases that you see here, 12:32.940 --> 12:35.740 this understanding the people you're designing for, 12:35.740 --> 12:38.700 this sort of ideation, brainstorming prototype, 12:38.700 --> 12:40.760 and then evaluating whatever it is 12:40.760 --> 12:42.100 that you've designed to see if it has 12:42.100 --> 12:44.760 the impact that you hoped it would. 12:45.540 --> 12:48.500 And, of course, human -centered design has been 12:48.500 --> 12:51.120 highly influential in the field of human-computer 12:51.120 --> 12:53.720 interaction. And now there's a tremendous amount 12:53.720 --> 12:56.960 for using it in other fields. Artificial intelligence, 12:56.960 --> 12:59.420 international development, public health. 13:01.620 --> 13:06.460 But as interest in human -centered design has grown, 13:06.460 --> 13:08.680 there are also these critiques of the approach that 13:08.680 --> 13:11.920 have emerged. And these critiques raise really important 13:11.920 --> 13:14.850 questions about whether the technological innovations 13:14.850 --> 13:17.740 that are frequently the outcomes of human-centered 13:17.740 --> 13:20.520 design, thinking about mobile phone applications, 13:20.670 --> 13:24.160 actually solve the problems that they are intended to. 13:24.740 --> 13:28.380 Scholars argue that the social change embodied in 13:28.380 --> 13:31.040 these innovations, such as improved access to health 13:31.040 --> 13:34.320 care or educational opportunities, rarely occur. 13:36.280 --> 13:40.260 building on this Dory Tunstall makes a very compelling 13:40.260 --> 13:43.500 case that this is a process that tends to benefit 13:43.500 --> 13:47.300 Western design and technology companies more so than the 13:47.300 --> 13:50.280 actual communities that participate in these projects 13:51.600 --> 13:56.140 finally Christina Harrington and her colleagues argue 13:56.140 --> 13:59.300 that human-centered design is an approach to developing 13:59.300 --> 14:03.040 that prioritizes developing novel technologies 14:03.040 --> 14:06.800 over deeply understanding the communities under study. 14:07.040 --> 14:10.280 They add that there's an opportunity for designers 14:10.280 --> 14:13.200 and technologists to expand their conversations 14:13.200 --> 14:15.840 about what outcomes are considered good 14:15.840 --> 14:18.280 design. And I think this is really important, and I 14:18.280 --> 14:20.700 want to say it again. They make a case that 14:20.700 --> 14:23.840 there's an opportunity for technologists and designers 14:23.840 --> 14:27.320 to broaden what they consider good design. 14:29.270 --> 14:30.680 So this brings me to what I'm 14:30.680 --> 14:31.780 going to talk about today. 14:32.040 --> 14:34.280 These critiques are valuable, and they have 14:34.280 --> 14:36.760 certainly forced me to do a tremendous amount of 14:36.760 --> 14:39.520 soul-searching about my own work, and I've 14:39.520 --> 14:42.060 concluded that this accompaniment approach can 14:42.060 --> 14:44.760 provide a meaningful response to these critiques. 14:45.250 --> 14:46.640 So what is it? 14:47.260 --> 14:50.200 Accompaniment started in liberation theology in 14:50.200 --> 14:53.300 Latin America, where faith groups learned to walk 14:53.300 --> 14:55.960 alongside the communities they were fighting for, for 14:55.960 --> 14:59.900 justice, instead of simply offering top-down help. 15:00.240 --> 15:04.020 Paul Farmer, a noted anthropologist and doctor, 15:04.020 --> 15:06.740 later brought the idea into global health and 15:06.740 --> 15:09.680 demonstrated how health care could only succeed if 15:09.680 --> 15:12.320 it was built with and sustained by communities. 15:13.220 --> 15:15.820 In research, accompaniment really isn't 15:15.820 --> 15:18.640 a method. It's more of a mindset, and you 15:18.640 --> 15:21.720 can use it to guide how you use methods. 15:23.160 --> 15:26.500 Use the methods to shape the relationships you build. 15:28.860 --> 15:31.180 the relationships you build, and how 15:31.180 --> 15:33.000 you determine success in research. 15:33.140 --> 15:36.580 And within technology design, it means shifting 15:36.580 --> 15:39.040 away from these quick fixes towards 15:39.040 --> 15:42.080 approaches that are deeply grounded in listening, 15:42.140 --> 15:44.240 mutual understanding, and 15:44.240 --> 15:46.480 working towards long-term impact. 15:47.060 --> 15:50.140 And I want to argue that it's an approach that can 15:50.140 --> 15:54.320 complement human-centered design research in HCI fields, 15:54.840 --> 15:57.460 Especially for the ways that it emphasizes, 15:57.460 --> 15:59.220 and again, these are ideas I'm very much 15:59.220 --> 16:02.140 borrowing from Carl DeSalvo, Carl's work. 16:02.140 --> 16:05.060 The way it forces you to relinquish control 16:05.200 --> 16:07.740 and turning the gaze on us. And by that, I 16:07.740 --> 16:10.200 mean HCI researchers, students, et cetera. 16:10.960 --> 16:13.080 So I'm gonna walk you through what 16:13.080 --> 16:14.980 this looks like in the rest of my talk. 16:16.560 --> 16:17.960 And this brings me to the 16:17.960 --> 16:19.520 first project I wanna discuss. 16:19.960 --> 16:24.240 Here we used human-centered design to work with Kenyan 16:24.240 --> 16:27.580 youth and their caregivers to develop an intervention 16:27.580 --> 16:31.660 to support their self -management of type 1 diabetes. 16:34.340 --> 16:37.640 This is an ongoing project that was originally funded 16:37.640 --> 16:40.760 by the National Institutes of Health. It's done in 16:40.760 --> 16:43.580 collaboration with faculty at Michigan State and in the 16:43.580 --> 16:48.300 spirit of accompaniment, done with an NGO in Nairobi 16:48.300 --> 16:51.200 called Rural Outreach Africa who I've worked with on 16:51.200 --> 16:54.580 a number of projects, as well as the Kenya Diabetes 16:54.580 --> 16:58.080 Management and Information Center, or DMI for short. 16:58.900 --> 17:05.260 This project was supported by an NIH program 17:05.260 --> 17:08.280 that was aimed at developing innovative mobile 17:08.280 --> 17:11.240 health interventions. And you can see this is sort 17:11.240 --> 17:13.100 of the information about the program and the 17:13.100 --> 17:17.040 project that you can find on the NIH's website. 17:17.280 --> 17:21.240 An interest in this topic has been motivated by 17:21.240 --> 17:26.310 increased access to mobile phones in Kenya and other 17:26.310 --> 17:29.760 low and middle income countries, primarily because 17:29.760 --> 17:32.160 of this perception that they can be used to 17:32.160 --> 17:35.080 help marginalized groups access health information. 17:35.680 --> 17:38.860 We did not develop a mobile phone app. 17:39.600 --> 17:42.580 Instead, by relinquishing control of what the 17:42.580 --> 17:45.220 NIH wanted, it supported a different, and I 17:45.220 --> 17:49.100 would argue a more compelling an unexpected 17:49.100 --> 17:52.900 outcome that includes this paper-based logbook 17:52.900 --> 17:54.940 or diary that you see in the picture here. 17:57.100 --> 17:59.000 So just a little background, 17:59.260 --> 18:01.460 why focus on type 1 diabetes? 18:01.580 --> 18:05.320 One reason is that people with 18:05.320 --> 18:07.320 type 1 diabetes is rising, 18:07.540 --> 18:12.340 and it continues to rise, especially among 18:12.340 --> 18:15.180 adolescents or people between the years of 10 and 18:15.180 --> 18:17.900 19, and especially in low- and middle-income 18:19.740 --> 18:21.860 countries. And I understand that a lot of the 18:21.860 --> 18:24.540 early work at the intersection of diabetes and 18:24.540 --> 18:26.740 technology has been done at Georgia Tech and 18:26.740 --> 18:29.540 continues to be done here. But just as a quick 18:29.540 --> 18:32.840 reminder, type 1 diabetes is a chronic condition 18:32.840 --> 18:35.100 where your body can't produce enough insulin. 18:35.440 --> 18:39.120 Managing type 1 of diabetes includes monitoring blood 18:39.120 --> 18:43.160 sugar levels, injecting insulin multiple times a 18:43.160 --> 18:47.760 day, adhering to a balanced diet and exercise routine, 18:47.760 --> 18:50.980 and regularly meeting with healthcare providers. 18:51.400 --> 18:54.680 Living with type 1 diabetes can be challenging 18:54.680 --> 18:57.700 anywhere, and I imagine that you can probably appreciate 18:57.700 --> 19:00.680 that it's especially challenging in countries like 19:00.680 --> 19:05.040 Kenya due to limited healthcare resources, economic 19:05.040 --> 19:07.940 barriers that make it difficult to access insulin 19:07.940 --> 19:10.620 and other supplies that you need, as well as what 19:10.620 --> 19:12.380 you could call these sort of cultural factors. 19:12.380 --> 19:15.280 There's a Sigma that surrounds having type 1 diabetes. 19:16.900 --> 19:20.540 A little bit more background, we focused on youth 19:20.540 --> 19:23.480 because this is a very important developmental stage 19:23.480 --> 19:26.480 for developing the habits that will help you live 19:26.480 --> 19:30.040 with diabetes throughout the course of your life. 19:30.460 --> 19:32.880 And again, in the spirit of an accompaniment, 19:32.880 --> 19:35.900 this is very much a long -term project between 19:35.900 --> 19:39.160 Michigan State University, DMI, as well as 19:39.160 --> 19:42.600 a really, really talented Kenyan design team. 19:42.780 --> 19:46.140 The project began in September of 2019. 19:46.530 --> 19:48.820 It was delayed for about a year and 19:48.820 --> 19:51.410 a half because of the COVID pandemic. 19:52.180 --> 19:56.000 But since then, over the course of the project, 19:56.000 --> 19:59.060 we've had five engagements in Kenya between the 19:59.060 --> 20:02.720 research design team and roughly 50 participants 20:03.070 --> 20:07.290 the last meeting was in June of last year and 20:07.880 --> 20:10.700 it's worth noting that this work has 20:10.700 --> 20:13.380 taken place in an urban and rural area 20:13.760 --> 20:17.420 the rural area in Vahiga which is close to Lake 20:17.420 --> 20:20.300 Victoria and it's important to recognize you 20:20.300 --> 20:22.420 know there are vast differences in terms of 20:22.420 --> 20:25.300 employment opportunities access to health care 20:25.300 --> 20:28.120 infrastructure between urban places like 20:28.120 --> 20:32.080 Nairobi and the more rural areas like in Vahiga. 20:34.160 --> 20:37.800 So we followed a very typical textbook human-centered 20:37.800 --> 20:40.760 design process. It was very qualitative in 20:40.760 --> 20:43.820 nature. I think qualitative research is especially 20:43.820 --> 20:47.080 conducive for developing relationships and research. 20:47.240 --> 20:50.120 Data collected over the course of the study 20:50.120 --> 20:53.760 primarily included interviews, pages and pages 20:53.760 --> 20:56.840 of field notes, as well as artifacts created 20:56.840 --> 20:59.940 by participants, design sketches, photographs. 21:00.440 --> 21:03.380 And as is typical in human-centered design, 21:03.380 --> 21:05.900 we began with this understanding or formative 21:05.900 --> 21:08.820 research phase with this ultimate goal of 21:08.820 --> 21:11.200 learning as much as we can about the people 21:11.200 --> 21:13.320 and the context that we're designing for. 21:14.860 --> 21:15.900 Here 21:20.650 --> 21:23.370 we use interviews and observations. We went to 21:23.370 --> 21:26.070 participants' homes to conduct these. Again, 21:26.070 --> 21:29.350 this is fairly typical in HCD. It's my colleague, 21:29.350 --> 21:31.290 Jenny Olson, who conducted the interviews. 21:31.530 --> 21:34.410 And by going to people's homes, it was not only useful 21:34.410 --> 21:37.550 for collecting people's data, but really respecting 21:37.550 --> 21:41.690 their lived realities and meeting them where they are. 21:42.570 --> 21:45.650 We asked them questions about their diagnosis, 21:45.690 --> 21:48.110 their everyday lives, their relationships, 21:48.150 --> 21:49.550 the food they ate. 21:49.550 --> 21:53.390 We learned about the improvised insulin storage 21:53.390 --> 21:55.990 units that they developed. Again, many of these 21:55.990 --> 21:58.970 households are off their country's electricity grid. 21:58.990 --> 22:01.450 Insulin needs to be kept cool and 22:01.450 --> 22:04.590 there's no limited access to refrigeration. 22:04.590 --> 22:07.010 We talked about using glucometers. 22:07.590 --> 22:11.470 And given the NIH's focus on mHealth, we asked 22:11.470 --> 22:14.030 people a lot of questions about their mobile phones, 22:14.030 --> 22:16.630 access, ownership, and what they use them for. 22:17.390 --> 22:20.770 So this data was really useful for collecting 22:20.770 --> 22:24.410 information about the users that we were designing for, 22:24.410 --> 22:27.110 but it also helps sort of lay the foundation 22:27.110 --> 22:30.090 between the research team and the participants that's 22:30.090 --> 22:32.870 really extended over the course of the project. 22:36.010 --> 22:39.890 Other methods that we use that I think it's safe 22:39.890 --> 22:42.930 to say are a little less typical in HCI and 22:42.930 --> 22:45.670 more popular in fields like public health are this 22:45.670 --> 22:48.230 photo study method. And again, it involved 22:48.230 --> 22:51.570 giving participants cameras, asking them to take 22:51.570 --> 22:54.070 photographs. We followed it up with interviews. 22:54.680 --> 22:58.230 And again, this is something that's very much aligned 22:58.230 --> 23:00.790 with this accompaniment approach because it's about 23:00.790 --> 23:03.730 giving participants, letting them tell their own 23:03.730 --> 23:07.570 stories and research and providing us with a deeper 23:07.570 --> 23:10.690 insight into their everyday lives that isn't always 23:10.690 --> 23:14.290 available with interviews and observations. Briefly 23:19.210 --> 23:20.890 talk about what we found, 23:21.250 --> 23:24.050 there's sort of a tension here. Again, unlike 23:24.050 --> 23:27.370 traditional approaches to research, accompaniment 23:27.370 --> 23:29.950 is really about emphasizing relationships 23:29.950 --> 23:32.950 over research findings, but it's hard to 23:32.950 --> 23:34.790 give a talk and not talk about your findings. 23:34.870 --> 23:37.610 And, you know, unsurprisingly, we found that 23:37.610 --> 23:40.410 managing type 1 diabetes in Kenya is hard. 23:41.190 --> 23:43.730 Insulin and supplies are often scarce, they're 23:43.730 --> 23:46.490 unaffordable. In rural areas, like I've 23:46.490 --> 23:49.470 mentioned, there's no reliable refrigeration, which 23:49.470 --> 23:51.990 makes it very difficult to keep insulin cool. 23:52.170 --> 23:55.100 Food insecurity makes it incredibly difficult 23:55.100 --> 23:58.670 to control your blood sugar levels. And 23:58.670 --> 24:00.820 again, there's a Sigma about the condition, 24:01.580 --> 24:02.890 misunderstandings that type 1 24:02.890 --> 24:05.710 diabetes is contagious, and it's not. 24:06.730 --> 24:09.810 And again, there's more information about this in the 24:09.810 --> 24:13.030 publication. But I think more importantly when 24:13.030 --> 24:16.410 analyzing this data when coding and listening to participants 24:16.410 --> 24:18.750 interviews looking at their photos and listening 24:18.750 --> 24:23.130 to their stories it was useful for deeply understanding 24:23.130 --> 24:26.450 their perspectives and recognizing that a mobile 24:26.450 --> 24:29.450 phone app was not going to address their challenges. 24:31.390 --> 24:34.310 And again in the spirit of accompaniment I'm gonna 24:34.310 --> 24:37.150 let one of the participants in our studies explain 24:37.150 --> 24:41.370 why this is the case. This is a short video from Anne. 24:41.370 --> 24:43.750 She has a son named Louis who's been managing type 24:43.750 --> 24:47.830 1 diabetes for 15 years now and she does a tremendous 24:47.830 --> 24:50.490 job of explaining the challenges with using a 24:50.490 --> 24:54.070 mobile device to support type 1 diabetes management. 24:54.210 --> 24:57.110 This is about four minutes long. 25:05.440 --> 25:06.580 Practice this. 25:09.760 --> 25:12.940 Tim I don't know I don't want to unplug this 25:12.940 --> 25:14.920 but I I don't know, this worked and we 25:14.920 --> 25:19.400 tested it for something funny going on now. 26:02.860 --> 26:06.860 There are cheaper providers like telecom, 26:06.940 --> 26:09.570 we could make a deal to solve that, but 26:09.570 --> 26:11.700 still it would be a problem to many people. 26:12.260 --> 26:14.200 The cost of a smartphone, 26:15.300 --> 26:17.940 for you to have an app, you will need a smartphone. 26:18.340 --> 26:20.340 You will need a couple of thousands 26:20.340 --> 26:22.520 of shillings to purchase one. 26:24.040 --> 26:28.300 So, maybe if we solve this issue, we could come up 26:28.300 --> 26:32.120 with the government to provide the spirals which is 26:32.120 --> 26:37.520 which is very ambitious but uh the the causes um an 26:37.520 --> 26:41.080 issue the other issue is um not everybody knows 26:41.080 --> 26:44.610 actually that one especially in the rural areas they 26:44.920 --> 26:48.940 only know how to pick they make a mother in all the 26:48.940 --> 26:55.880 jobs and i don't know she can just pick it um so the 26:55.880 --> 26:58.900 the literacy because a bit of it could be an issue, 26:59.570 --> 27:03.490 but it could be solved with training and maybe 27:04.030 --> 27:10.500 figuring around network in remote areas is problematic. 27:10.700 --> 27:12.600 There are some places where you go, you know, 27:14.240 --> 27:17.200 nobody will get you on board unless now 27:17.200 --> 27:19.310 you start changing a network, another network, 27:19.310 --> 27:21.480 but still you're not available full time. 27:21.860 --> 27:24.240 then accessibility in schools. 27:24.500 --> 27:26.360 Our issue is the youth. 27:26.520 --> 27:29.280 Like my son is in Syria, but how will he use his 27:29.280 --> 27:33.560 school? So it could be an issue, but then our group 27:33.560 --> 27:39.020 suggested that if your youth has an analytic issue, 27:39.340 --> 27:43.190 then we need to strike or to have a policy that 27:43.190 --> 27:45.890 they can leave the phone with the nurse so 27:45.920 --> 27:48.580 that we can be monitoring the app. And you will 27:50.420 --> 27:53.200 So there are a number of things that are going on 27:53.200 --> 27:55.200 in this video. You know, Anne clearly knows what 27:55.200 --> 27:57.600 people need, you know, free access to smartphones. 27:58.240 --> 28:00.720 She also does a wonderful job of 28:00.720 --> 28:03.110 articulating all of the challenges with mobile 28:03.110 --> 28:05.040 phones, not being able to use them. 28:05.080 --> 28:08.520 You know, and Mike introduced me by saying I sort 28:08.520 --> 28:10.900 of have a negative take on these technological 28:10.900 --> 28:13.840 interventions, but it's worth noting that everything 28:13.840 --> 28:16.830 that Anne is describing has more or less persisted, 28:16.830 --> 28:19.280 particularly in rural areas, for the past 15 28:19.280 --> 28:22.100 years that I've studied mobile phones in rural parts 28:22.100 --> 28:25.240 of Kenya. But, you know, coming back to my talk, 28:25.460 --> 28:28.820 mHealth efforts are predicated on this idea that you 28:28.820 --> 28:31.160 own a mobile phone and that you're able to consistently 28:31.160 --> 28:33.800 use it. You know, that's clearly not the case in 28:33.800 --> 28:37.140 Kenya, and Ann does an excellent job of detailing why. 28:38.100 --> 28:41.220 So after completing this formative research, it 28:41.220 --> 28:43.780 really became clear that we had to change our 28:43.780 --> 28:47.980 direction and in terms of accompanying accompaniment 28:47.980 --> 28:51.880 it meant relinquishing control to let go of the 28:51.880 --> 28:55.280 expectations of how the project might turn out or how 28:55.280 --> 28:58.460 we spelled it out in our application to the NIH 28:58.640 --> 29:02.120 it gave us an opportunity to use these grant 29:02.120 --> 29:06.100 resources in very different ways and by letting go of 29:06.100 --> 29:08.880 this anticipated outcome it provided space for 29:08.880 --> 29:11.740 what I think are more impactful outcomes to emerge. 29:14.090 --> 29:17.320 So to arrive at these outcomes and again this 29:17.320 --> 29:19.620 is very typical in human centered design we 29:19.620 --> 29:23.760 held we engaged in this ideation phase where we 29:23.760 --> 29:27.500 generated creative ideas and potential solutions 29:27.500 --> 29:30.560 that centered on users needs and insights. 29:30.900 --> 29:33.720 We continued to collaborate with the same youth 29:33.720 --> 29:36.440 and caregivers that were in our interview phase of 29:36.440 --> 29:39.580 the project and at this point the project we also 29:39.580 --> 29:42.700 involved our very talented Kenyan design team. 29:43.340 --> 29:46.640 So we held two day-and -a-half long workshops, 29:48.290 --> 29:51.580 hotels in Vejiga and in Nairobi, and of 29:51.580 --> 29:53.900 course we paid for participants transportation, 29:54.040 --> 29:55.640 provided them with per diems 29:55.640 --> 29:57.700 and meals during these sessions. 29:58.900 --> 30:02.240 So I want to return to this photo study method and 30:02.240 --> 30:05.160 how it was especially aligned with accompaniment in 30:05.160 --> 30:07.480 the way that it seemed to build community and shared 30:07.480 --> 30:10.640 understanding among everyone involved in our project. 30:11.160 --> 30:14.800 Both design workshops followed a similar protocol. 30:14.880 --> 30:18.180 After introductions, we asked participants to 30:18.180 --> 30:20.800 explain their photos from the earlier phase, 30:20.800 --> 30:23.300 and you can see some of those photos here. 30:28.860 --> 30:31.580 So when we were at the workshops, we asked the youth 30:31.580 --> 30:34.480 and their caregivers to talk about their pictures, 30:34.740 --> 30:37.550 they talked about using their glucometers, 30:37.860 --> 30:40.240 injecting insulin, eating food, and 30:40.240 --> 30:42.680 exercising. all of the activities that 30:42.680 --> 30:44.560 were really crucial to their well-being. 30:44.640 --> 30:47.260 They shared selfies and reflected on some of the 30:47.260 --> 30:50.400 more challenging aspects of managing type 1 diabetes, 30:50.400 --> 30:52.580 such as the pain that comes with injecting 30:52.580 --> 30:55.860 yourself with insulin every day, as well as some of 30:55.860 --> 30:58.540 the more positive moments of, you know, successfully 30:58.540 --> 31:01.380 managing their blood sugar levels or simply 31:01.380 --> 31:03.720 living with the condition for a number of years. 31:03.900 --> 31:07.160 But this activity of sharing and discussing the 31:07.160 --> 31:10.070 photos was useful for a couple of things. for 31:10.070 --> 31:12.360 confirming a number of the findings from the earlier 31:12.360 --> 31:15.600 phases of our research and more significantly 31:15.600 --> 31:19.000 motivating discussions among our participants, 31:19.000 --> 31:21.640 the youth and their parents. And they talked about 31:21.640 --> 31:24.140 their shared experience and realized they 31:24.140 --> 31:27.460 weren't the only people managing type one diabetes. 31:27.800 --> 31:31.180 It sort of built solidarity and a sense of community, 31:31.440 --> 31:34.760 especially among the youth and caregivers, but 31:34.760 --> 31:37.140 also among the research and design team that 31:37.140 --> 31:40.000 were also in attendance at these workshops. I 31:40.000 --> 31:42.120 think an important thing to keep in mind is that 31:42.120 --> 31:44.740 most of these participants had never met anybody 31:44.740 --> 31:48.280 else with type 1 diabetes before our project. 31:48.500 --> 31:52.160 So this trust and some intimacy seemed 31:52.160 --> 31:54.360 to support the project's next phases. 31:57.280 --> 32:00.280 And that's where we asked participants to generate 32:00.280 --> 32:03.500 a collection of ideas. We asked participants 32:03.500 --> 32:07.480 to sketch their ideas, to help support diabetes 32:07.480 --> 32:10.480 self-management, This gave them an active 32:10.480 --> 32:13.340 role in shaping the design rather than researchers 32:13.340 --> 32:15.560 coming in and telling them what they need. 32:15.780 --> 32:19.160 I love drawing. I think it's a wonderful 32:19.160 --> 32:22.000 way to make abstract ideas tangible. 32:22.000 --> 32:25.400 It helps participants express their perspectives in 32:25.400 --> 32:29.200 ways that might be hard to put in words. It also fosters 32:29.200 --> 32:34.430 a shared sense of ownership around ideas and strengthens 32:34.560 --> 32:41.260 both voice and community in the design process. 32:41.260 --> 32:44.700 And you can see a few of those drawings here. 32:47.600 --> 32:51.060 So there was one concept that appeared over and over 32:51.060 --> 32:54.100 in both the rural and urban workshops. And it was 32:54.100 --> 32:58.360 this idea for a diary or a logbook. And when discussing 32:58.360 --> 33:01.300 this concept, groups consistently talked about how 33:01.300 --> 33:04.200 it could simultaneously address a number of needs. 33:04.400 --> 33:07.060 It could be used to remind them of their doctor's 33:07.060 --> 33:10.900 appointments, to document the foods they ate and how much 33:10.900 --> 33:14.060 insulin they were injecting. It could also be used 33:14.060 --> 33:17.140 as a tool to help others learn about type 1 diabetes. 33:17.760 --> 33:21.060 And unlike an mHealth app, it didn't require 33:21.060 --> 33:25.320 owning a smartphone or having internet access. this 33:32.380 --> 33:34.200 idea of relinquishing control. 33:34.220 --> 33:38.220 So at this point in the project, 33:38.240 --> 33:39.840 our Kenyan design team and 33:39.840 --> 33:41.840 collaborators really ran with things. 33:42.360 --> 33:45.360 After over a three month period, they took the 33:45.360 --> 33:48.020 sketches that had emerged from our workshop, 33:48.360 --> 33:51.260 turned them into three initial versions of the 33:51.260 --> 33:54.260 diary, and then developed a final prototype 33:54.260 --> 33:56.700 of the diary that you see here on the screen. 33:57.580 --> 34:00.180 And all of this is documented in a paper 34:00.180 --> 34:01.600 if you want to learn more about it. 34:02.520 --> 34:05.540 So the final phase of the human-centered design 34:05.540 --> 34:08.540 process, the sort of implementation or formative 34:08.540 --> 34:12.840 evaluation phase. We did this in two parts. The 34:12.840 --> 34:15.820 first involved bringing study participants back 34:15.820 --> 34:18.960 together for another workshops at the same hotel. 34:21.840 --> 34:24.600 So I mean I'm getting ahead of myself. So we 34:24.600 --> 34:27.040 did the evaluation in two phases. The first 34:27.040 --> 34:30.600 was a more conventional formative evaluation 34:30.600 --> 34:33.120 that you see in human -centered design where 34:33.120 --> 34:36.180 we evaluated the diary with participants. 34:36.180 --> 34:39.240 And the second one was a little less common. We brought 34:39.240 --> 34:42.080 participants back together again to really ask them 34:42.080 --> 34:45.260 to sort of evaluate and reflect on their experience of 34:45.260 --> 34:48.240 participating in this human -centered design process. 34:50.820 --> 34:53.280 So again, there's this tension here. 34:53.280 --> 34:56.600 I really wanna emphasize how accompaniment 34:56.600 --> 34:59.280 is an approach to research that prioritizes 34:59.280 --> 35:01.680 relationships over reporting findings. 35:01.800 --> 35:04.500 At the same time, I feel very much compelled 35:04.500 --> 35:06.920 to share sort of the outcomes of our research. 35:07.320 --> 35:10.440 So we asked participants to engage in this 35:10.440 --> 35:12.820 sort of usability test with the prototype. 35:12.980 --> 35:15.640 We gave them the diary to use over a 35:15.640 --> 35:18.340 month long period. We did a pre and post 35:18.340 --> 35:20.480 survey to see what happened with it. 35:21.600 --> 35:24.920 After analyzing the data, we generally found that 35:24.920 --> 35:27.820 participants' reaction to the diary was positive. 35:27.820 --> 35:31.800 They commented on the diary's attractive appearance. 35:31.800 --> 35:34.920 You can see here there is a real appreciation 35:34.920 --> 35:38.280 for having illustrations of foods that they ate. 35:38.280 --> 35:41.280 Also information that was very important to their 35:41.280 --> 35:44.280 context, such as developing these improvised 35:44.280 --> 35:48.460 insulin storage units using clay pots and charcoal. 35:48.990 --> 35:51.200 And there was evidence from our surveys 35:51.200 --> 35:53.280 that our participants did use it 35:53.280 --> 35:55.030 every day over a month-long period. 35:55.280 --> 35:58.520 We also gained a tremendous amount of information 35:58.520 --> 36:01.580 on how to improve the design of the diary. And 36:01.580 --> 36:03.800 so there was a general sense that this was a feasible 36:03.800 --> 36:06.580 idea and of course we clearly need to conduct 36:06.580 --> 36:09.260 more research to see what if any outcome this 36:09.260 --> 36:13.300 has on health outcomes, what impact it might have. 36:16.900 --> 36:20.620 So like I said we also returned to Kenya in April 36:20.620 --> 36:24.860 of 2023 to sort of evaluate participants experiences 36:24.860 --> 36:27.980 in the project and this is where my understanding 36:27.980 --> 36:31.140 of these other outcomes really began to emerge. 36:32.100 --> 36:34.980 So seven months following our evaluation of the diary 36:34.980 --> 36:38.720 many things happened over that time period and you 36:38.720 --> 36:42.180 can see that in the picture here. We presented the 36:42.180 --> 36:46.540 final version of our diary to the final version of our 36:46.540 --> 36:50.160 diary was presented to our participants. This is an 36:50.160 --> 36:53.740 event held in Vahiga County. You can see the the county 36:53.740 --> 36:57.540 governor and ministry of health was at this event. 36:57.540 --> 37:00.420 We held other events. This is an event called a 37:00.420 --> 37:05.240 Heroes Lunch, where type 1 diabetes champions or people 37:05.240 --> 37:09.230 who advocate and support others with type 1 diabetes 37:09.230 --> 37:12.540 came together. You can see that with the cake here. 37:13.640 --> 37:17.000 Finally, our collaborator, Eric Omonde, was 37:17.000 --> 37:20.220 featured on a local TV station where he talked 37:20.220 --> 37:22.600 about the project and the impact it had. 37:22.700 --> 37:24.820 And again, this seemed useful for sort of 37:25.140 --> 37:28.760 educating people about type 1 diabetes. 37:29.200 --> 37:32.120 And in terms of evaluation, it was these 37:32.120 --> 37:35.400 experiences which seemed most impactful, more 37:35.400 --> 37:38.010 so than developing the actual diary. And 37:38.060 --> 37:41.050 again, more relevant to my broader argument, 37:41.580 --> 37:44.960 by prioritizing these outcomes which were not 37:44.960 --> 37:47.900 necessarily the focus of our original project. 37:50.760 --> 37:54.700 So when returning to evaluating participants' 37:54.700 --> 37:58.320 experiences in the workshop, we were evaluating 37:58.320 --> 38:00.500 participants' experience in the project. 38:04.320 --> 38:07.180 We We asked participants to reflect on 38:07.180 --> 38:09.920 their experience and write them down 38:09.920 --> 38:11.940 on the comment cards that you see here. 38:11.940 --> 38:14.040 And again, it was these experiences 38:14.040 --> 38:15.960 that were most impactful. 38:15.960 --> 38:18.560 And also some surprising ones. 38:18.560 --> 38:20.500 Participants frequently talked about the 38:20.500 --> 38:22.120 camera that we gave them in the early 38:22.120 --> 38:24.700 stages of the study, and many of them continued 38:24.700 --> 38:26.900 to use it for all kinds of purposes. 38:27.680 --> 38:31.220 The caregivers, who were overwhelmingly women, 38:31.320 --> 38:33.880 consistently talked about enjoying 38:33.880 --> 38:35.660 being able to stay in a hotel, 38:35.800 --> 38:38.900 having somebody else cook their meals. There was 38:38.900 --> 38:41.960 sort of a sense that some dignity came with being 38:41.960 --> 38:45.200 in this project and being included in the research. 38:45.300 --> 38:48.240 And finally, the most consistent thing that 38:48.240 --> 38:50.460 came out of the workshop was just making friends. 38:51.500 --> 38:53.900 Community building among the youth with 38:53.900 --> 38:57.000 type 1 diabetes seemed to emerge as 38:57.000 --> 38:59.400 a significant outcome of our project. 38:59.720 --> 39:02.060 Participants have created WhatsApp groups and 39:02.060 --> 39:05.820 exchanged personal experiences, they continued to support 39:05.820 --> 39:09.180 each other after the workshops ended and many of 39:09.180 --> 39:11.540 them described again this is the most significant 39:11.540 --> 39:21.360 outcome of the project so accompaniment does call 39:21.360 --> 39:25.440 for this long term sustained relationships and this 39:25.440 --> 39:29.180 past June we met with our participants again this 39:29.180 --> 39:31.720 is roughly three years following the beginning of 39:31.720 --> 39:35.320 the project. We brought participants back together, 39:35.820 --> 39:37.780 continued to explore their 39:37.850 --> 39:40.320 experiences being in the project, 39:40.610 --> 39:43.680 also had conversations with them about how to sustain 39:43.680 --> 39:46.560 the project, and it's been really exciting to watch 39:46.560 --> 39:50.340 them grow up. Many of them are in college now, and again 39:50.340 --> 39:52.760 our collaborators continued to sort of be running 39:52.760 --> 39:55.680 with the project. A few months ago I learned that they 39:55.680 --> 39:58.600 had received money from Novo Nordisk or in the process 39:58.600 --> 40:02.280 of putting 20,000 copies of the diary to distribute. 40:02.740 --> 40:05.140 In terms of my own work I've submitted a 40:05.140 --> 40:07.440 proposal to the NIH to more systematically 40:07.440 --> 40:10.360 evaluate the impact of the diary but it's 40:10.360 --> 40:12.020 unclear what's going to happen with that. 40:15.250 --> 40:18.040 So just quickly before I discuss you know sort of 40:18.040 --> 40:20.420 the takeaways from this research I wanted to show 40:20.420 --> 40:23.660 another image of the diary. I think it's a really 40:23.660 --> 40:26.920 beautiful well-designed artifact that speaks to 40:26.920 --> 40:29.840 how human-centered design can work and really is a 40:29.840 --> 40:32.800 credit to our deeply talented Kenyan design team. 40:35.220 --> 40:38.060 Okay, so sort of what are the takeaways from this? 40:38.480 --> 40:41.540 Coming back to this broader idea of accompaniment 40:41.540 --> 40:45.120 and relinquishing control, I deeply believe 40:45.120 --> 40:48.400 it allows for us to have a more authentic human 40:48.400 --> 40:52.910 -centered design process, one that embraces ambiguity 40:52.910 --> 40:55.980 and not assuming that a technical intervention 40:55.980 --> 40:59.200 is the right outcome. In a lot of ways, I 40:59.200 --> 41:01.700 think the NIH and the reviewers got things wrong. 41:02.180 --> 41:05.420 Strictly speaking, you can't anticipate what's going 41:05.420 --> 41:08.400 to happen in human-centered design, so it seemed 41:08.400 --> 41:11.340 a little off to have this baked into our project that 41:11.340 --> 41:14.340 a mobile phone application would be the outcome. 41:14.720 --> 41:16.980 And again, you know, emphasizing 41:16.980 --> 41:18.680 that this design process, 41:18.680 --> 41:20.980 using it as a way to learn about each other 41:20.980 --> 41:24.020 and support relationships, can be more 41:24.020 --> 41:26.820 impactful and will likely be much more long 41:26.820 --> 41:29.160 -lasting than the technological intervention. 41:29.470 --> 41:31.860 I think this makes a strong case for why 41:31.860 --> 41:34.780 we can need to continue to integrate more 41:34.780 --> 41:38.120 visual methods into HCI and ICTD research. 41:38.480 --> 41:41.240 And finally, and this last one can be a 41:41.240 --> 41:43.380 little hard to swallow as a researcher, 41:44.040 --> 41:48.020 accompaniment demands flexibility and adaptation and 41:48.020 --> 41:51.360 putting aside your personal interests and ambitions. 41:51.570 --> 41:53.890 So I don't know if I'll be able to evaluate 41:53.890 --> 41:56.280 this diary but recognizing that there's a 41:56.280 --> 41:58.480 tremendous amount of success for me stepping 41:58.480 --> 42:00.820 away from the project and giving it to our 42:00.820 --> 42:03.880 local to our Kenyan collaborators to run with. 42:06.140 --> 42:09.780 So at the time that I have left, what brings me to the 42:09.780 --> 42:12.900 second project I want to discuss and it's the development 42:12.900 --> 42:16.560 implementation of this technology design course. 42:17.060 --> 42:19.660 And here I want to demonstrate how 42:19.660 --> 42:23.200 accompaniment can shape how we teach HCI. 42:29.030 --> 42:31.270 This is a project that's been funded by 42:31.270 --> 42:34.130 the Mozilla Foundation, in particular 42:34.130 --> 42:36.550 their Responsible Computing Challenge. 42:36.690 --> 42:38.990 This is a program that supports the 42:38.990 --> 42:43.150 conceptualization, development, and piloting of curricula 42:43.150 --> 42:45.190 that empowers students to think about the 42:45.190 --> 42:48.190 social and political context of computing. 42:48.370 --> 42:51.470 It is a fantastic program. I can't speak 42:51.470 --> 42:54.080 highly enough about working with the Mozilla 42:54.080 --> 42:55.930 Foundation. And I know that they've 42:55.930 --> 42:58.010 supported some projects here at Georgia Tech. 42:58.010 --> 43:01.310 And here I want to demonstrate how accompaniment 43:01.310 --> 43:05.250 can really upend how we teach HCI and ICTD. 43:05.810 --> 43:09.130 Courses that are generally meant to train computer 43:09.130 --> 43:12.950 and information science students to design technologies 43:12.950 --> 43:15.750 that solve problems in developing countries. 43:16.350 --> 43:19.230 And this is an approach that reproduces 43:19.230 --> 43:21.850 this narrative that Africans are technologically 43:21.850 --> 43:25.110 backwards, lack innovation, and only 43:25.110 --> 43:27.630 use technology instead of developing it. 43:27.690 --> 43:30.770 And in my proposal, I suggested that we 43:30.770 --> 43:33.970 do this by incorporating topics from 43:33.970 --> 43:36.610 the field of African studies into HCI. 43:36.610 --> 43:38.610 So again, this is basically a course 43:38.610 --> 43:41.210 where you have HCI plus African studies. 43:41.210 --> 43:44.070 And African studies is a really exciting 43:44.070 --> 43:46.290 discipline that's sort of an umbrella for 43:46.290 --> 43:48.490 all of these topics that are important. 43:49.090 --> 43:51.710 Colonialism, multilingualism, 43:51.890 --> 43:53.330 resource constraints, 43:53.330 --> 43:55.090 gender, race, etc. 43:57.750 --> 44:00.410 So there are many things that motivated this 44:00.410 --> 44:03.610 idea. I think the most influential are my 44:03.610 --> 44:06.690 experiences teaching a week-long human-centered 44:06.690 --> 44:09.390 design course at Kenya's Egerton University. 44:09.390 --> 44:12.490 I've done this for the last four summers now. 44:12.950 --> 44:15.730 Michigan State and Egerton have worked together 44:15.730 --> 44:19.310 for nearly 50 years. They're both agricultural 44:19.310 --> 44:21.890 institutions. There are a number of research 44:21.890 --> 44:24.830 and exchange programs between the universities. 44:25.450 --> 44:30.110 And what stands out most from these experiences is how 44:30.110 --> 44:32.930 creative and how thoughtful the students' technology 44:32.930 --> 44:36.810 design projects are. They consider its topics like 44:36.810 --> 44:42.170 sustainability, privacy, misinformation, access, because 44:42.170 --> 44:45.310 these are issues that deeply affect them in ways 44:45.310 --> 44:50.130 that seem a little less obvious to the students who are 44:50.130 --> 44:52.930 largely socialized in the United States. You know I 44:52.930 --> 44:55.250 teach a similar version of this human-centered design 44:55.250 --> 44:59.390 course and I'm constantly reminded of how much can be 44:59.390 --> 45:01.530 learned from these students and their perspectives. 45:03.770 --> 45:06.530 So again this idea of incorporating these perspectives 45:06.530 --> 45:08.930 into an HCI course was really central to 45:08.930 --> 45:13.010 this idea. There are other ways that I did this. I'm 45:13.010 --> 45:16.390 clearly not African. I co-taught the course with 45:16.390 --> 45:19.270 an African studies scholar. This is Dr. Chodi, 45:19.270 --> 45:21.790 who is a linguistics professor at Michigan State. 45:22.130 --> 45:25.210 We assigned readings that were written by African 45:25.210 --> 45:28.690 scholars about technology and design, in particular 45:28.690 --> 45:31.470 ones that sought to sort of change this narrative 45:31.470 --> 45:34.310 that Africa is sort of the recipient of technology 45:34.310 --> 45:37.790 designed in Silicon Valley, but that there's a 45:37.790 --> 45:40.090 lot of innovation taking place in the continent. 45:40.390 --> 45:44.250 We invited the authors from these, authors of 45:44.250 --> 45:47.550 these texts to speak in our class and have conversations 45:47.550 --> 45:49.850 with students about the assigned readings 45:49.850 --> 45:52.990 and see my colleague Dr. Komen from Daystar 45:52.990 --> 45:56.330 University sort of joining our class. 45:56.450 --> 45:58.370 And then we engage in this 45:58.370 --> 46:00.130 human centered design project. 46:00.130 --> 46:03.670 Students identified a research challenge, 46:03.670 --> 46:04.990 conducted research, 46:05.730 --> 46:08.970 brainstorm ideas and then prototyped it see a paper 46:08.970 --> 46:11.910 prototype there and again all of this was done in 46:11.910 --> 46:16.800 collaboration with students at Egerton University and 46:16.800 --> 46:19.950 what I'm most proud of and I think one of the central 46:19.950 --> 46:22.730 features of this idea was the development of a 46:22.730 --> 46:26.610 program I called technologists in residence so we had 46:26.610 --> 46:29.390 five students and one faculty member from Kenya's 46:29.390 --> 46:32.870 Egerton University they were selected through a competitive 46:32.870 --> 46:36.190 process they received a stipend and laptop to 46:36.190 --> 46:39.750 really recognize their expertise and their contributions 46:39.750 --> 46:43.290 to this class. They participated in the class 46:43.290 --> 46:47.330 online for most of the semester via Zoom and then they 46:47.330 --> 46:50.110 traveled to Michigan and see us all at the Detroit 46:50.110 --> 46:54.030 Airport there and these students acted as mentors you 46:54.030 --> 46:56.610 know they weren't sort of passive participants in 46:56.610 --> 46:59.830 the class but they took a very active role in mentoring 46:59.830 --> 47:01.910 the Michigan State students over the course of 47:01.910 --> 47:05.130 the semester as they worked on their design projects. 47:06.630 --> 47:09.430 Just quickly, the first iteration of 47:09.430 --> 47:11.610 this course, highly interdisciplinary, 47:12.430 --> 47:15.490 we teach it one day a week for two and a 47:15.490 --> 47:18.270 half hours which has benefits and drawbacks. 47:18.610 --> 47:20.770 Navigating the eight -hour time difference is 47:20.770 --> 47:23.330 always a challenge, but a lot of the sort 47:23.330 --> 47:26.410 of findings that I talked about here come 47:26.410 --> 47:28.670 from build notes that I took following each 47:28.670 --> 47:32.130 class and there's a paper about this as well. 47:34.690 --> 47:37.290 So again the goal was sort of to upend 47:37.290 --> 47:40.430 these traditional ICT course and challenge 47:40.430 --> 47:43.330 this idea that we need to design technology 47:43.610 --> 47:46.850 interventions to address problems in other parts of 47:46.850 --> 47:50.510 the world. And a lot of the class was devoted to simply 47:50.510 --> 47:54.310 talking to one another and introducing sort of 47:54.310 --> 47:57.850 everyday life in Kenya to the Michigan State students. 47:58.150 --> 48:01.110 One of the most simple, straightforward 48:01.110 --> 48:03.510 and effective ways I did this was by 48:03.510 --> 48:06.250 developing this day -in-the-life exercise. 48:07.110 --> 48:10.790 Kenyan students created PowerPoint presentations that 48:10.790 --> 48:14.670 walked the MSU students through a typical day from the 48:14.670 --> 48:16.950 time they wake up in the morning to the time they go 48:16.950 --> 48:20.210 to bed at night. Their presentations included lots of 48:20.210 --> 48:24.030 pictures. Their stories that detailed their everyday 48:24.030 --> 48:29.630 routines, what they ate, how they prepared their food, 48:29.750 --> 48:33.690 sort of the multilingual realities of their life, 48:33.690 --> 48:37.550 using English in the classroom, Swahili on campus, 48:37.550 --> 48:40.490 and then their mother tongue language at home. 48:41.210 --> 48:44.110 And I mean, many of us have been teachers in 48:44.110 --> 48:47.270 a classroom. There are very few instances 48:47.270 --> 48:50.370 where I've just observed students be so engaged 48:50.370 --> 48:52.930 and so captivated by what they're watching on 48:52.930 --> 48:55.510 the screen than during these presentations. 48:56.320 --> 48:58.930 So, of course, they found contrast between 48:58.930 --> 49:01.150 the Kenyan students and their own lives. 49:01.150 --> 49:03.910 Most aren't speaking multiple language. 49:03.910 --> 49:06.190 There's a tendency to eat more processed 49:06.190 --> 49:08.390 food in the United States than in Kenya. 49:08.470 --> 49:11.220 But they also found a lot of commonalities, 49:11.350 --> 49:13.830 you know, a love for playing video games, 49:13.910 --> 49:18.470 navigating how to work a job and balance school. 49:18.750 --> 49:22.830 So these and other activities seem successful in sort 49:22.830 --> 49:26.410 of breaking down the stereotype as African as other. 49:31.630 --> 49:34.550 For having conversations over the course of the 49:34.550 --> 49:38.710 semester and created space for some vulnerability 49:38.710 --> 49:42.890 and openness to discuss topics that can sometimes 49:42.890 --> 49:45.510 be difficult to talk about in the classroom. 49:45.630 --> 49:47.390 We talked a great deal about 49:47.390 --> 49:49.070 the impacts of colonialism. 49:49.510 --> 49:51.510 Kenyan students talked about how this 49:51.510 --> 49:54.150 continued to impact their lives. They talked 49:54.150 --> 49:57.690 about naming practices and how they don't 49:57.690 --> 50:00.050 typically go by the names of Jim and Paul. 50:00.250 --> 50:02.150 They use these names instead 50:02.150 --> 50:03.590 of their indigenous names. 50:03.750 --> 50:07.270 They described how their parents' and grandparents' 50:07.270 --> 50:11.310 land was taken during British colonial rule. There were 50:11.310 --> 50:14.950 lots of discussions about the differences in cost of 50:14.950 --> 50:18.190 living between Kenya and the United States. Of course, 50:18.190 --> 50:21.090 in Kenya, things tend to be a lot less expensive. 50:21.550 --> 50:24.590 And again, relevant to technology, we had guest 50:24.660 --> 50:28.750 speakers who talked about infidelity and plural marriage 50:28.750 --> 50:31.290 and how this shaped women's mobile phone use. 50:31.650 --> 50:35.070 And these conversations were also reciprocal. 50:35.110 --> 50:38.550 It created a space for the Kenyan students to ask 50:38.550 --> 50:41.270 questions to the Michigan State students. And we 50:41.270 --> 50:44.110 devoted a significant amount of class time to 50:44.110 --> 50:47.110 just explaining how what the Kenyan students sometimes 50:47.110 --> 50:50.750 see on television and in movies aren't necessarily 50:50.750 --> 50:53.850 reflective of reality in the United States. 50:56.970 --> 50:59.550 So these and other discussions also 50:59.550 --> 51:02.130 affected the students design projects. 51:03.090 --> 51:06.670 Classes were devoted to engaging in critiquing of 51:06.670 --> 51:09.010 ideas where the Kenyan students played significant 51:09.010 --> 51:12.730 role. When Michigan State students proposed something 51:12.730 --> 51:15.490 like a mobile phone app you know they quickly said 51:15.490 --> 51:18.150 most people in rural areas don't have smartphones. 51:18.390 --> 51:20.730 They talked about some of the challenges 51:20.730 --> 51:23.630 with accessing the internet in rural areas. 51:23.630 --> 51:26.310 Lots of discussions over what languages to 51:26.310 --> 51:28.690 use and the implications of those choices. 51:32.340 --> 51:35.110 So again this brings me to another aspect of 51:35.110 --> 51:37.670 accompaniment that's relevant to how we design 51:37.670 --> 51:41.610 interventions in HCI and ICTD. It seemed 51:41.610 --> 51:45.270 like the course was useful for sort of turning 51:45.270 --> 51:48.090 the gaze on us, and what do I mean by that? 51:48.750 --> 51:51.650 It seemed like through the course the students 51:51.650 --> 51:54.250 working together as partners, they had these repeated 51:54.250 --> 51:57.430 engagements with the Kenyan students via Zoom 51:57.430 --> 52:01.590 and also via WhatsApp. It helped them sort of realize 52:01.590 --> 52:04.330 these weren't people to study or to be designed 52:04.330 --> 52:08.730 for, but seemed to push them to first reflect on 52:08.730 --> 52:10.750 their own positionality. I mean, many of these 52:10.750 --> 52:14.510 students have socialized in the American Midwest and 52:14.510 --> 52:18.680 to deeply reconsider where expertise and technological 52:18.680 --> 52:20.670 innovation comes from. Again, these are two 52:20.670 --> 52:22.910 things I really hope to achieve with the course. 52:23.060 --> 52:25.330 They really seem to move away from this 52:25.330 --> 52:27.670 mindset of helping someone or studying 52:27.670 --> 52:30.530 someone and learn how it might be more effective 52:30.530 --> 52:32.390 to actually design and work together. 52:35.370 --> 52:37.650 So again, similar to the findings of my other 52:37.650 --> 52:40.250 projects, I'm not going to spend too much time 52:40.250 --> 52:42.970 going over all the outcomes. I have a student 52:42.970 --> 52:45.790 who's developed a really beautiful website that 52:45.790 --> 52:49.290 details every aspect of the class, including 52:49.290 --> 52:51.810 the projects, the assignments, the syllabus. 52:52.450 --> 52:54.670 If you teach an HCI class, I 52:54.670 --> 52:56.070 encourage you to look at this. 52:59.320 --> 53:02.030 and again the takeaways here I think there are a couple 53:02.030 --> 53:06.690 um again incorporating this accompaniment mindset 53:06.690 --> 53:10.490 into the class I think was successful in sort of 53:10.490 --> 53:15.070 challenging how we more traditionally teach HCI and ICTD 53:15.070 --> 53:18.390 the course seems successful of course you know there 53:18.390 --> 53:21.650 are opportunities to continue evaluating it but I 53:21.650 --> 53:23.570 think one of the most important things that I've taken 53:23.570 --> 53:26.470 away from it is that you know Africa is typically 53:26.470 --> 53:29.230 thought of sort of peripheral or maybe you have one 53:29.230 --> 53:31.470 class to it or you have a separate class it's about 53:31.470 --> 53:34.230 technology use in developing countries but I think 53:34.230 --> 53:36.430 there are a lot of opportunities that might come from 53:36.430 --> 53:39.730 making Africa more central to how we teach human computer 53:39.730 --> 53:43.510 interaction and at least provides opportunities 53:43.510 --> 53:46.590 for different types of outcomes you know not just a 53:46.590 --> 53:50.570 piece of technology for your portfolio but learning 53:50.570 --> 53:53.910 how to engage in cross -cultural dialogue and sort of 53:53.910 --> 53:56.830 this global understanding of technology use that I 53:56.830 --> 53:59.590 think is increasingly important for designers to have. 54:00.070 --> 54:02.990 I also want to add that, of course, some aspects 54:02.990 --> 54:05.910 of doing this project demand resources, particularly, 54:05.910 --> 54:08.050 you know, bringing the Kenyan students to the 54:08.050 --> 54:10.950 United States. But a lot of the things that we did 54:10.950 --> 54:13.570 are pretty accessible and I think there's a case to 54:13.570 --> 54:16.810 be made for integrating these into more courses. 54:20.040 --> 54:23.390 so I'm beginning to wrap things up so you know one 54:23.390 --> 54:25.910 of the limitations of accompaniment is how do you 54:25.910 --> 54:29.570 sustain it and that's very much where I am now last year 54:29.570 --> 54:31.870 around this time I had a pretty significant grant 54:31.870 --> 54:34.570 from the U.S. State Department it provided an opportunity 54:34.570 --> 54:37.750 to really grow this program grow the technologists 54:37.750 --> 54:40.630 and residents program have an exchange program that 54:40.630 --> 54:43.530 would also take Michigan State students to travel 54:43.530 --> 54:46.110 to Kenya to evaluate their projects we were going to 54:46.110 --> 54:49.510 have a big project showcase at the embassy in Nairobi. 54:49.510 --> 54:52.230 And then almost a year ago to the day, I got a 54:52.230 --> 54:55.330 letter from Marco Rubio saying that this was not aligned 54:55.330 --> 54:58.370 with the government's priorities. And they clawed 54:58.370 --> 55:00.730 back every last cent of money that they had given to 55:00.730 --> 55:03.190 Michigan State and that the project came to a halt. 55:03.650 --> 55:05.710 This is one of many projects at 55:05.710 --> 55:07.910 Michigan State that's lost funding. 55:08.690 --> 55:12.430 And, you know, it has multiple types of impacts, 55:12.430 --> 55:15.170 including really eroding the trust that I've Spent 55:15.170 --> 55:17.530 years developing with my colleagues at Egerton 55:17.530 --> 55:21.250 University because it meant we couldn't pay them. 55:21.690 --> 55:24.070 More broadly, these cuts have created a lot 55:24.070 --> 55:27.970 of certainty about the future of funding, the 55:27.970 --> 55:30.790 type of research that's really important to 55:30.790 --> 55:33.330 helping me sustain the work that I like to do. 55:35.380 --> 55:38.000 But if anything, I mean, this is a reminder 55:38.000 --> 55:41.310 that this work is more important now than ever. 55:41.890 --> 55:43.930 Encouraging students to think critically 55:43.930 --> 55:47.550 about power, who defines problems, who controls 55:47.550 --> 55:50.290 resources, and whose knowledge counts. 55:50.640 --> 55:52.610 So that's one thing that I think is an 55:52.610 --> 55:54.890 important thing to take away from this talk. 55:55.030 --> 55:57.550 A couple of others before I conclude. 55:57.550 --> 56:00.870 Just to remind you of what motivated all this, 56:00.870 --> 56:03.570 these critiques of human -centered design, the 56:03.570 --> 56:07.750 impacts of these typically technological interventions, 56:07.750 --> 56:11.230 who benefits from them. and again, a need 56:11.230 --> 56:14.110 to really expand our conversations over what is 56:14.110 --> 56:17.190 good design. And I hope that I've achieved this. 56:17.350 --> 56:21.090 But the most important thing that I'm reflecting on is 56:21.090 --> 56:24.570 that by incorporating accompaniment in how that we do 56:24.570 --> 56:28.750 our work, it seems to make for a more human process, 56:28.870 --> 56:32.730 one that's very slow and relational and open-ended. 56:32.730 --> 56:35.470 And I do think now more than ever, there's a 56:35.470 --> 56:38.170 urgency of bringing that way of thinking into 56:38.170 --> 56:40.360 how we design and teach about technology. 56:41.230 --> 56:44.130 And just to say this again, I deeply 56:44.130 --> 56:45.690 believe that it leads to a more 56:45.690 --> 56:48.030 authentic human -centered design process. 56:50.910 --> 56:53.490 So I'll leave you with that, and that concludes 56:53.490 --> 56:56.910 my talk. I have the names of the people who've 56:56.910 --> 57:00.450 helped me do this research. I definitely want to 57:00.450 --> 57:02.990 thank the Mozilla Foundation for paying for me to 57:02.990 --> 57:05.150 come here and talk about this, and thank you for 57:05.150 --> 57:07.430 your time. And I'd be happy to take questions. 57:09.710 --> 57:11.630 Thank you so much.